z Stenhouse, G.
er heterogeneity
dian Journal of
n, W., Hall, F,,
rest disturbance
note Sensing of
'olfe, R., Hall, F.
Landsat surface
90—2000. IEEE
72.
lysis in Sakarya
GIS, Msc. ITU
ects of irrigation
ion, Geography
. and Demir, H.,
and water use in
Implications for
an. 20: 467-488.
). The effect of
r the status of
, 61, 210-220.
5 S, Kumar, V.,
ents in terrestrial
ata sets. Global
jg X. Tu 1, €&
ie change in the
ional Journal of
nney, M. P, &
change detection
rect atmospheric
0—244.
nce study. GAP
o. 15 (Revised).
t Venture. Tokyo
"Water Projects"
gy , Stockholm,
R. Larson, B. K.
jn of the 1990's
erminous United
., 67, 650-662.
isek, J. G., Irons,
nuity: Issues and
mote Sensing of
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XXXIX-B8, 2012
XXII ISPRS Congress, 25 August — 01 September 2012, Melbourne, Australia
SUBDIVISION OF PANTANAL QUATERNARY WETLANDS: MODIS NDVI TIME-
SERIES IN THE INDIRECT DETECTION OF SEDIMENTS GRANULOMETRY
N. C. Penatti *' & T. I. R. de Almeida *
? Institute of Geoscience, University of Sáo Paulo, Rua do Lago, 562 - Sáo Paulo - Brasil
(natasha.penatti, talmeida)@usp.br
Commission VIII, WG VIII/6
KEY WORDS: Vegetation, MODIS, Geology, Comparison, Environment, Multitemporal, TIMESAT
ABSTRACT:
The Pantanal is the world's largest wetland presenting a variety of ecological sub-regions. The region is characterized by seasonal
floods followed by long droughts. At this period, some areas rapidly dry, while others remain soaked. The study hypothesis was
based on the statement that this phenomenon cannot be totally explained by small relief variations but by the sediment granulometry:
the pelitic sediments allow the soil to retain moisture longer, implying that the vegetation has greater possibilities of continuing
photosynthetically active even during the drought. It was developed based on the spectral behaviour of MODIS products, validated
by previous fieldwork. Using MODIS, we studied a large scale patterns in spatial and seasonal dynamics of the vegetation in
different regions of Pantanal. So, two indirect parameters of the local physical environment — sediment granulometry and water
availability — potentially can be estimated. We calculated the NDVI from MODO9GQ for rainy and dry seasons, generating triplets
(NDVI/NIR and Red bands) that allowed to identify vegetation changes in those periods. The 16-days composites of NDVI
(MOD13Q1) were used to generate a 5-year time-series for pixels associated with 161 locals sampled for granulometric analyses.
The samples were taken in 10 different areas from the 20 geological and environmentally homologous areas delimited in this
research. The clear tendency in the time-series confirms the working hypothesis, indicating that there is a high relationship between
drought-related changes in vegetation extracted from NDVI and sediment texture, parameter that plays an important role in soil
moisture, influencing the vegetation response to droughts.
1. INTRODUCTION
The Pantanal is a vast wetland predominantly covered by
savannas and, more locally, by forests. It is located in a tropical
region at the center of South America over a sedimentary basin.
This basin, originated at the end of Tertiary, is yet tectonically
active, due to efforts between the tectonic plates of South
America and Nazca (Assumpgdo, 1998). Nested in the Upper
Paraguay River basin, the floodplain is located mostly in Brazil
with smaller areas in Bolivia and Paraguay. Pantanal is a
complex wetland characterized by seasonal floods with varied
intensities, and markedly wet and dry phases, with geomorphic
features derived from the present conditions and others
inherited from successive Pleistocene and Holocene climate
changes. This landscape has enormous internal variability due
largely to the geological environment: the nature of the
sediments, small relief variations associated with the
sedimentary or erosional processes and neotectonics.
The large size of this feature and the access difficulties makes
the remote sensing an indispensable tool. To detect
phenological changes, the remote sensing community has relied
on vegetation indices, usually derived from the Advanced Very
High Resolution Radiometer (AVHRR) or the Moderate
Resolution Imaging Spectroradiometer (MODIS) as these
sensors have the advantages of greater image acquisition
frequency and larger areal coverage (Redo and Millington,
2011). MODIS products are particularly suitable to study
Pantanal because they can cover the entire wetland in a single
orbit, providing a synoptic view of this dynamic region. From
these data it is possible to extract products of vegetation indices
which enhance the spectral response of plants, reducing the
influence of soil, aiming to distinguish the vegetation
phenology and changes in land use/land cover at the analysis of
a temporal image series (Wardlow et al., 2007).
Currently, satellite derived vegetation index data, such as
Normalized Difference Vegetation Index (NDVI), plays an
important role for vegetation drought response and flood
monitoring (Kogan, 1998; McVicar and Jupp 1998, Ji and
Peters 2003, Gu et al., 2008, Caccamo et al., 2011). The indices
are useful for detection and monitoring large area vegetation
stress resulted from drought or soil oversaturation following
flooding and excessive rains. In dynamic areas, such Pantanal,
annual cycle of floods and droughts, changes the landscape
completely, creating different inside areas based on physical
environmental characteristics. Between November and March,
rains saturate the area; rivers overflow and flood the lowland
forest and grassland areas, turning the plains into an immense
freshwater wetland. From April onwards though, the waters
start to recede, and by October, water can only be found in low-
lying ponds and rivers. Although, drought effects in vegetation
have a tendency to be more persistent in some areas than in
others. The study hypothesis was based on the statement that
this phenomenon cannot be totally explained by the small relief
variations in Pantanal but by the sediment granulometries. In
this context, the main objective of this study was to understand
how MODIS-based NDVI correlates to sediments granulometry
over the seasons and along a five year NDVI time-series in
Pantanal. The knowledge of the velocity and intensity of NDVI
values decrease at the sampled areas will enable the
establishment of granulometric patterns for regions not visited
during the fieldwork. This will provide an unpublished and
essential indicative granulometric spatial data, that can allow
the proposal of a subdivision of the Pantanal in geological and
environmental homologous sub-areas.
